A small test POC connecting books using Open Linked Data ==> start adding values... Today we see a lot of initiatives but when we check LIBRIS XL a book it is not linked and in RAÄ a book it is linked to a church in Wikipedia but no good user interface exist and its just a link or one link per object... compare Wikidata Q61765464 for the same book that also links the book to people mentioned in the book....
I see old organisations having problem using new technology as grafdatabases and having an user community with the wrong skill matrix i.e. we get Metadatadebt and the price we pay is data that is not FINDABLE/USABLE...
In Sweden cultural people like Digisam has been speaking about Linked data since 2012 but I feel they miss a vision and dont have the right skill matrix. This is a POC what can be done if we move the Riksantikvarieämbetet UGC data to a modern platform with the possibilities to easier connect those books as sources for a grave.... we have the same problem with the National Library project LIBRISXL that also have the books in RDF but as "Strings not things" see picture below were "Riksantikvarieämbetet" is a string and not an object like Wikidata Q631844 i.e. #METADATADEBT...
Example how a book can be connected to people mentioned in the book on a modern platform like Wikidata with > 6800 properties
- Wikidata Phabricator Task T215603 Connect WD Churches and church documentation at RAÄ
- Video in Swedish describes this linked solution where I extract 200 books from LIBRIS and combines them with RAÄ and Wikidata. A small reaction on this tweet from LIBRIS XL that I feel has no Linked Data vision displayed in what is delivered <-> LIBRISXL tweet "UX-design – att forma en användarupplevelse"
- Map with the churches that are documented in a RAÄ Book (also links LIBRIS/LIBRIS XL/sv:Wikipedia)
Access LIBRIS and search for Svenska Kyrkor and extract items available from RAÄ Samla
- Search in LIBRIS is ZSER:(Sveriges kyrkor)
1-1) Check if RAÄ SAMLA has references to Wikipedia/Wikidata using UGC
1-2) export data
- Next step use Open Refine to Match items in Wikidata to find Objects for authors etc.. and upload to Wikidata
2-1) Upload to Wikidata --> SAMLA books will be one object with references from the WD church object and same as with the authors of the books
Python get_libris_svenska_kyrkor.py --> log
- Output
- svenskakyrkor.csv
- contains LIBRIS information looks like the JSON doesnt contain page information see unanswered question
- samlaWikidata.csv
- contains UGC relations Samla <-> Wikidata
- every book is an object in WD
- example Wikidata objekt Q61723424 is RAÄ Samla 6880 - see also What links here
- every book has one or more Main Topic P921 that is the churches they describe or if a grave is described the buried person is linked
- every church gets a described by P1343 linking back to the book
- Map with churches described by RAÄ
- Linked data Graph of book Riddarholmskyrkan same as
- RAÄ Samla kulturarvsdata.se/raa/samla/6873
- LIBRIS 19512784
- Map - chuches near me documented by RAÄ
- Start with a book about the Riddarholm church see tweet book Wikidata Q61765464
- Wikidata/Sälgö
- Wikidata
- Task Wikidata Phabricator Task T215603 and workboard
- Category:P1260_Error
- Github
- Wikidata
- RAÄ/UGC
- No formal error reporting is used at RAÄ UGC so I sent a tweet to Abbe that I have a WD category with odd things Category:P1260_Error - I guess either I update WD or they do redirect feels like some Link rot
- LIBRIS XL
- has a discussion group that is answered but lack a change process that you can follow
- you get no error id
- there is no public backlog
- you get no notification when something is in production you have to ask every 6 months in the discussion group example status federated search
- has a discussion group that is answered but lack a change process that you can follow
- See if we can do this for indoor maps T217470
- Question LIBRIS XL about searching link
- LIBRIS XL product owner Anna Berggren reply 20190214
- JSON API is just available for LIBRIS and not LIBRIS XL
- Riksantikvarieämbetet is a text string rdfs:Literal and not the correct linked data way owl:sameAs for more info about ProvisionActivity see LIBRIS Data Model
- its confirmed that pattern matching in URLS and labels is what we have today(2019-feb) and use old school LIBRIS API
- Feels an odd approach to have such bad data when LIBRIS have invested in Linked Data building LIBRIS XL since 2012/roadmap 2014. LIBRIS statement about Open Data
- UGC-hubben (användargenererat innehåll)
- Sveriges kyrkor - Riksantikvarieämbetes Samla
- LIBRIS Xsearch
- Wikidata/Commons contribution strategies for GLAM organizations
- Petscan query find people in Wikipedia category Kategori:Gravsatta_i_Riddarholmskyrkan and not source wd:Q61765464
Ett försök att få kopplingar våra kyrkogårdar med kyrka, flygbilder, Find A Grave, Släktforskarförbundet se github:salgo60/Gravstensinventeringen-Wikidata